The stem cell theory of cancer predicates that both normal and cancer stem cells can be induced into or released from dormancy depending on their multipotential constraints and microenvironment restraints. A unified theory of cancer predicts that even though genetic makeup in cancer dormancy maybe pivotal, cellular context must be paramount. After all, both normal stem cells and cancer stem cells proliferate and differentiate. They can be stationary and migratory. They can be static and dynamic. Importantly, gonadal germ cells are prototype stem cells and germ cell tumor of the testis (TGCT) is a model stem cell cancer.
1. Seed and Soil
In 1889, Stephen Paget published that breast cancer metastasized to certain organ sites (e.g., liver) far more often than others (e.g., spleen)
[1]. To account for this salient observation, he proposed the classic seed and soil hypothesis. Importantly, this seminal hypothesis may explain another key cancer hallmark besides metastasis, namely dormancy. In fact, dormancy is innately ingrained in a seed, and is intimately influenced by the soil. A seed can remain dormant forever. It is similar to a normal or cancer stem cell that can renew and resume in perpetuity. The very being of a seed is dependent on the soil. This is the very essence of a stem cell theory of cancer.
In many respects, the idea if not the observation of dormancy contradicts a genetic theory of cancer. Although genetic mutations may be obligatory in the initiation and promotion of cancer, they may not contribute to the pathogenesis of cancer dormancy. After all, genetic mutations are less likely to occur and unlikely to be selected for in the absence of any cellular growth, division, or activity
[2]. Since dormancy is an intrinsic property of stem cells (and certain somatic cells), it may not be necessary to invoke any specific mutations for those cells to be induced into or released from dormancy. The discovery that cells with identical genetic background can display variable functional phenotypes, including dormancy
[3], and cells with certain integrin profiles have unique dormancy potential without regard to any specific genetic defects
[4] is also in conflict with the genetic theory of cancer.
To be dormant or not depends on the nature and status of both the cell and its microenvironment
[5]. Paracrine factors at the site of metastasis, such as bone morphogenetic proteins, may induce dormancy in metastasis-initiating cells by inhibiting their capacity to self-renew
[6]. Conversely, metastasis-initiating cells (similar to their adult stem cell counterparts) secrete Coco (an antagonist of transforming growth factor (TGF)-beta ligands), which reactivates dormant breast cancer cells in the lung
[7]. Similarly, periostin, another matrix protein secreted by stromal fibroblasts in response to TGF-beta in stem cell niches, promotes activation from dormancy by facilitating the presentation of Wnt ligands to tumor cells
[8]. Thus, dormancy reaffirms a fundamental principle of cancer: the malignant phenotype can be both dynamic and static and is intricately dependent on a close interplay between the involved cell and its niche.
2. Stemness
When stemness is an overt characteristic of cancer, it could be the code to decipher a conundrum of cancer dormancy. In many respects, stemness accounts for all the conventional hallmarks of cancer—autonomy, metastasis, heterogeneity, immune evasion, and genetic instability
[9][10]. When stemness is a preeminent feature of cancer, it alludes to a stem cell origin and stem-like nature of cancer.
Similar to other cancer hallmarks, dormancy is ingrained in stemness, and stemness is embedded within dormancy. A prime example of a dormant stem cell is an inactive germ cell. Without fertilization, an ovum may remain dormant for the rest of the host’s life span. After fertilization, it will form all the diverse organs and tissues in an offspring. It is of interest that transformed fetal gonocytes in the form of intratubular germ cell neoplasia unclassified (IGCNU)
[11] remain dormant and do not become malignant until after puberty a decade or several decades later.
There is irony in stemness and dormancy. Stemness is good because regeneration and wound healing is good. However, stemness is also bad when it is prevalent in malignancy. Dormancy is good when a cancer is in slumber, but it is also bad because the cancer may wake up. Hence, a putative dormancy factor, TGF-beta, is both a tumor suppressor, because it mediates anti-proliferative and apoptotic effects, and a tumor promoter, because it induces tumor motility, metastasis, and EMT. TGF-beta is a different actor in a fetus vs. an adult, during embryogenesis vs. carcinogenesis. What it does in the fetus may be perfectly benign during embryogenesis, but the same activity in an adult may be patently malignant during carcinogenesis.
In many respects, stemness is an epiphany of oncology recapitulating ontogeny. The exact role and proper function of a normal stemness or a malignant stem-like factor is dependent on the exact or proper context, i.e., timing and placement. If cancer has a stem cell origin and is a stem cell disease, it does not need to reprogram or reinvent its stemness. It does not need to hijack or retrieve what it already owns.
3. Primary vs. Metastatic Cancer
A stem cell origin of cancer can better account for dormancy and metastasis in the seed and soil for primary and metastatic tumors than a genetic origin of cancer does. After all, once a primary tumor forms, there are not many more driver mutations that appear and accumulate
[12]. A multipotent progenitor cell is more capable of spreading to and colonizing various sites than an oligopotent or unipotent progenitor cell. They provide and receive different sets of cues from their respective microenvironments to become exuberant or remain dormant.
The observation that primary tumors and metastatic lesions are different diseases and can be modified by specific microenvironmental factors has far-reaching biological implications and clinical applications. For example, the 21-gene recurrence score (Oncotype Dx) predicts risk of local breast cancer recurrence (ipsilateral breast, chest wall, and regional nodal) regardless of systemic disease control with tamoxifen or chemotherapy
[13]. After induction therapy, the maintenance and prolongation of remission by targeting minimal residual disease (i.e., cancer stem cells) and the bone microenvironment (e.g., with anti-inflammatory and anti-metabolic agents) may improve clinical outcome regardless of local disease control in the past or in the future
[14].
Importantly, when a primary tumor or its metastatic lesions are inherently stable or dormant, surgical extirpation with curative intent of the former may be feasible and with improved control of the latter may be appropriate. Turajlic et al.
[15] demonstrated that primary renal cell carcinoma (RCC) with low intratumoral heterogeneity (ITH), but elevated somatic copy-number alterations (SCNAs) had rapid progression at multiple sites. However, those with low ITH and low fraction of the genome affected by SCNAs had overall low metastatic potential. Results from their study suggest that the removal of primary tumor is beneficial for those indolent RCC with high ITH (even in the presence of metastasis) but not for those fulminant RCC with low ITH and high SCNAs (despite absence of metastasis).
Interestingly, low ITH high SCNAs reflect aneuploidy and implicate aberrant asymmetric division, suggesting that a stemness origin may be involved in the evolution of this malignant phenotype
[16].
4. Prolonged Remissions
Sometimes, disease outliers provide invaluable clues about cancer dormancy. Exceptional cases of extreme clinical outcomes (both good and bad) may unveil fundamental mechanisms of action in cancer dormancy.
When an active, threatening cancer calms down and becomes dormant, something must be keeping the molten lava bottled up, as in a dormant volcano, or the husky body bundled up, as in a hibernating bear. What physical barriers delay a volcano from erupting? What sleeping potions enable the bear to keep slumbering?
Prolonged remission is particularly intriguing in a supposedly deadly metastatic cancer when researcher do not expect durable remission or anticipate extended survival. This is not talking about patients with indolent tumors, such as follicular lymphomas or chronic lymphocytic leukemias, who may live for years if not decades without or despite treatment.
For example, Greenberg et al. showed that about 20% of patients with metastatic breast cancer who had experienced a complete remission after treatment would live beyond ten years, perhaps even twenty years
[17]. Surprisingly, even some patients with an incomplete remission could experience a prolonged remission. It seems that a small proportion of patients with other metastatic cancers, such as renal cell carcinoma and melanoma, may also attain prolonged remissions
[18][19].
A stem cell theory of cancer is consistent with and predicts the occurrence of prolonged remissions. It is conceivable that in any minimal residual disease after induction systemic treatments, the drug-resistant, non-cycling cells comprise cancer stem cells. It is plausible that when not triggering or rattle those putative cancer stem cells, they will remain quiescent, if not dormant, for a prolonged period, if not permanently.
5. Very Late Recurrence
The unique occurrence of very late recurrence is particularly noteworthy in the context of a stem cell theory of cancer
[20]. In most instances, the primary tumor has been extirpated. Therefore, the recurrent disease is likely related to micrometastases that eventually become reactivated and insurrected.
Approximately 1% of patients with germ cell tumor of the testis (TGCT) develop very late recurrence more than 5 years after diagnosis
[21]. The question is: What is unique about this 1% of TGCT that causes very late recurrence, and what can be done to prevent it?
Moore et al.
[22] examined a group of 25 TGCT patients with very late recurrence. The median time for relapse was 16.1 years. The longest time was 33.1 years. All recurrent tumors comprised somatically transformed tumor, yolk sac tumor, and/or teratoma.
The irony of cancer dormancy and very late recurrence is that what were considered favorable tumor phenotypes before the era of chemotherapy—indolent TGCT such as yolk sac tumor and teratoma—have become unfavorable entities since the inception of chemotherapy (in the 1970s). In 1946, about 90% of patients with metastatic TGCT died within one year of diagnosis
[23]. Today, over 90% of patients with the same diagnosis are cured
[24]. Unfortunately, many of the patients who still die from TGCT harbor indolent chemotherapy-resistant TGCT, such as yolk sac tumor and teratoma
[25][26].
In many respects, the indolent and dormant nature of yolk sac tumor and teratoma belies their treacherous eventual outcome. Importantly, ensconced within a residual, supposedly innocuous teratoma after chemotherapy are dormant and covert progenitor stem cells that may reawaken and relapse as a yolk sac tumor and/or a somatically transformed tumor
[27].
There is a reason certain teratomas are malignant and need to be removed to be cured. It is not the teratoma per se that is malignant. It is the dormant, aberrant progenitor stem cells lodged within the teratoma that make it malignant and necessary to be removed in a timely manner.
A stem cell theory of TGCT may be able to prevent very late recurrence and enable the remaining 10% of TGCT patients to be cured who still die from this otherwise very curable cancer.